An artificial neural network (ANN) for archaeometric studies was created to facilitate provenance attribution of archaeological ceramics. A multilayer perceptron model (MLP) was applied to construct the network, including only one hidden layer. Moreover, correction parameters based on historical and archaeological evidences were applied to Bayesian probability factor. The ANN was trained by using clays mixings mathematically constructed based on a reference chemical database of Sicilian sediments. The clay mixing takes in consideration compositional variability within the same geological site and the extent of the ceramic manufacture processes. Test was performed by querying the ANN with compositional data of ceramics found in archaeological sites coherent with clays sampling areas. Up to 88% correct attribution was verified, with good correspondence between geological and archaeological contexts. Finally, merits of ANN were highlighted by comparing the extent of successfully provisional attribution with classical statistical methods (PCA and LDA).

Artificial neural network for the provenance study of archaeological ceramics using clay sediment database

Raneri, Simona
2019-01-01

Abstract

An artificial neural network (ANN) for archaeometric studies was created to facilitate provenance attribution of archaeological ceramics. A multilayer perceptron model (MLP) was applied to construct the network, including only one hidden layer. Moreover, correction parameters based on historical and archaeological evidences were applied to Bayesian probability factor. The ANN was trained by using clays mixings mathematically constructed based on a reference chemical database of Sicilian sediments. The clay mixing takes in consideration compositional variability within the same geological site and the extent of the ceramic manufacture processes. Test was performed by querying the ANN with compositional data of ceramics found in archaeological sites coherent with clays sampling areas. Up to 88% correct attribution was verified, with good correspondence between geological and archaeological contexts. Finally, merits of ANN were highlighted by comparing the extent of successfully provisional attribution with classical statistical methods (PCA and LDA).
2019
Barone, Germana; Mazzoleni, Paolo; Spagnolo, Grazia Vera; Raneri, Simona
File in questo prodotto:
File Dimensione Formato  
ANN_clays_revisedmanuscript_21012019.pdf

accesso aperto

Tipologia: Documento in Post-print
Licenza: Creative commons
Dimensione 1.16 MB
Formato Adobe PDF
1.16 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11568/976869
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 26
  • ???jsp.display-item.citation.isi??? 23
social impact